Semiconductor memories generally include a multitude of memory cells arranged in rows and columns. Each memory cell is capable of storing digital information in the form of a “1” or a “0” bit. To write (i.e., store) a bit into a memory cell, a binary memory address having portions identifying the cell's row (the “row address”) and column (the “column address”) is provided to addressing circuitry in the semiconductor memory to activate the cell, and the bit is then supplied to the cell. Similarly, to read (i.e., retrieve) a bit from a memory cell, the cell is again activated using the cell's memory address, and the bit is then output from the cell.
Semiconductor memories are typically tested after they are fabricated to determine if they contain any failing memory cells (i.e., cells to which bits cannot be dependably written or from which bits cannot be dependably read). Generally, when a semiconductor memory is found to contain failing memory cells, an attempt is made to repair the memory by replacing the failing memory cells with redundant memory cells provided in redundant rows or columns in the memory.
Conventionally, when a redundant row is used to repair a semiconductor memory containing a failing memory cell, the failing cell's row address is permanently stored (typically in predecoded form) on a chip on which the semiconductor memory is fabricated by programming a nonvolatile element (e.g., a group of fuses, antifuses, or FLASH memory cells) on the chip. Then, during normal operation of the semiconductor memory, if the memory's addressing circuitry receives a memory address including a row address that corresponds to the row address stored on the chip, redundant circuitry in the memory causes a redundant memory cell in the redundant row to be accessed instead of the memory cell identified by the received memory address. Since every memory cell in the failing cell's row has the same row address, every cell in the failing cell's row, both operative and failing, is replaced by a redundant memory cell in the redundant row.
Similarly, when a redundant column is used to repair the semiconductor memory, the failing cell's column address is permanently stored (typically in predecoded form) on the chip by programming a nonvolatile element on the chip. Then, during normal operation of the semiconductor memory, if the memory's addressing circuitry receives a memory address including a column address that corresponds to the column address stored on the chip, redundant circuitry in the memory causes a redundant memory cell in the redundant column to be accessed instead of the memory cell identified by the received memory address. Since every memory cell in the failing cell's column has the same column address, every cell in the failing cell's column, both operative and failing, is replaced by a redundant memory cell in the redundant column.
Thus, for example, as shown in FIG. 1, a semiconductor memory 20 having failing memory cells 22, 24, 26, 28, 30, 32, 34, and 36 is repaired in the conventional manner described above using redundant rows 38, 40, and 42 and redundant columns 44, 46, and 48. As described above, the semiconductor memory 20 is repaired by replacing all memory cells in columns 50, 52, and 54, including failing memory cells 22, 24, 26, and 28, with redundant memory cells in redundant columns 44, 46, and 48. Further repairs to the semiconductor memory 20 are accomplished by replacing all memory cells in rows 56, 58, and 60, including failing memory cells 30, 32, 34, and 36, with redundant memory cells in redundant rows 38, 40, and 42. The process described above for repairing a semiconductor memory using redundant rows and columns is well known in the art, and is described in various forms in U.S. Pat. Nos. 4,459,685, 4,601,019, 5,422,850, and 5,528,539.
Unfortunately, it is difficult to provide enough redundant rows or columns in a semiconductor memory to repair all failing memory cells therein using the conventional repair process described above without using an excessive amount of space (commonly known as “real estate”) in the memory for the redundant rows or columns. With the increasing size of semiconductor memories continuously increasing the need for redundancy, memory designers find themselves caught between providing sufficient redundancy to successfully repair most memories. As a result, memory designers are either using excessive space in the memories or providing insufficient redundancy to save space in the memories and, as a result, having to discard memories that are unrepairable. Obviously, neither alternative is desirable.
U.S. Pat. No. 5,548,225 to Rountree et al. discloses a repair system that, in contrast to the conventional repair system described above, does not use an entire redundant row or column to repair each defective memory cell in a semiconductor memory. In the Rountree repair system, the column address of a defective memory cell is stored using fuses in the same manner as described above. In addition, a partial row address common to a group of cells in the defective cell's column that includes the defective cell itself is also stored using fuses. When a memory address having column and row addresses that match the stored column address and stored partial row address is received, a redundant memory cell in a spare column is accessed. As a result, all of the cells in the group identified by the stored column address and stored partial row address are replaced by redundant cells in the spare column, while those cells in the defective cell's column not in the identified group are not replaced. Thus, the efficiency of repairs is increased by the Rountree system because only some of the redundant cells in the spare column are used to repair the defective cell, while other redundant cells remain in the spare column to repair other defective cells.
Unfortunately, the Rountree repair system can be problematic as well, because storing a full column address and a partial row address for every defective memory cell in need of repair requires a great deal of storage space (e.g., fuses, etc.). Consequently, the ever-increasing size of modern semiconductor memories, and the corresponding increase in the number of defective memory cells typically found, makes the Rountree repair system increasingly prohibitive to use because of the amount of storage space it requires.
Therefore, there is a need in the art for an improved device and method for repairing a semiconductor memory containing a failing memory cell. Such a device and method should replace the failing cell with a redundant memory cell without replacing the failing cell's entire row or column with the redundant cell's entire row or column. The device should also replace multiple failing cells in different rows or columns with redundant memory cells in a single redundant row or column in order to make more efficient use of redundant rows and columns, and should do so without the excessive need for storage space characteristic of the Rountree repair system.